RIMBANet: Reconstructing Integrative Molecular Bayesian Networks
RIMBANet is a software package for reconstructing integrative molecular Bayesian networks. There are multiple sources of perturbations (eg. genetic mutations, copy number variations, methylations and etc.) that may contribute aberrant behaviors of biological systems such as cancer cells. Cells employ multiple levels of regulation that enable them to respond to genetic and environmental perturbations. At the transcriptional level, abundance of mRNA can be affected by the rate of transcription, a complex process regulated by transcription factors and enhancers, and by the rate of degradation of transcripts, a process regulated by RNA binding proteins and, in many organisms, microRNAs. Protein abundances are determined by protein degradation and protein synthesis rates, where protein synthesis can be regulated by translation initiation factors and microRNAs. Protein activity depends on a number of factors in addition to protein abundance, including protein localization, phosphorylation states and other post-translational modifications, and protein-protein interactions. In addition to transcript and protein levels, the abundance of small-molecule metabolites is also tuned in response to changes in a cell’s physiological state.
One of major goals of systems biology is to understand how these genetic and environment variations drive transcriptional networks, protein-protein interaction networks, metabolite networks and etc. to give arise to complex phenotypes. The integration of genetic variation and intermediate observations such as mRNA variations into probabilistic causal models that can dissect genetic pathways and provide mechanisms connecting DNA to clinical outcomes. We developed a computation framework centered around Bayesian network and implemented it in RIMBANet (BN4Distribution.tgz), which is freely available for download. We have previously used RIMBANet to discover causal relationships in complex human diseases such as diabetes and obesity and yeast model. Detail description of the RIMBANet package will be published in BMC journal Open Network Biology.
We applied RIMBANet to investigate how genetic variations regulate transcriptional and metabolite level changes in yeast. The full data set used in the study is available here (Yeast_4_Distribution.tgz).